Infracordal Stenosis: A Glucocorticoid-Responsive Subtype of Autoimmune Laryngotracheal Stenosis

Abstract

Objective:

As our understanding of autoimmune laryngotracheal stenosis (LTS) evolves, distinguishing patients who may benefit from systemic immunosuppression versus those needing only local treatment is increasingly important. In this study, we identify a distinct subset of autoimmune LTS characterized by edema of the inferior true vocal folds that extends to the superior aspect of the cricoid cartilage, termed “infracordal stenosis.” The objective of this study is to characterize the clinical presentation and treatment outcomes of infracordal stenosis and compare it to typical autoimmune-related subglottic stenosis (AI-SGS).

Methods:

We conducted a retrospective review of patients with autoimmune laryngotracheal stenosis evaluated by both rheumatology and otolaryngology at our institution to identify two groups: patients with infracordal stenosis and those with typical AI-SGS. Data on immunosuppressive treatments and airway dilation procedures were collected. Time to first dilation was compared between groups.

Results:

Among 49 patients with autoimmune LTS, 11 had infracordal involvement. Six patients had isolated infracordal stenosis while five had concomitant subglottic involvement. Kaplan-Meier analysis showed longer time to first dilation in patients with infracordal involvement (median 792 vs. 44 days; p = 0.048). Four out of six patients with isolated infracordal stenosis required no dilations during their entire follow-up period.

Conclusion:

Among autoimmune LTS patients referred to rheumatology, those with infracordal involvement experienced longer time to first dilation compared to those with typical AI-SGS. These findings suggest that infracordal stenosis may represent a distinct, glucocorticoid-responsive phenotype within autoimmune laryngotracheal stenosis, with implications for treatment selection and multidisciplinary care.

Introduction

Laryngotracheal stenosis (LTS) is characterized by narrowing of the airway at the level of the cricoid cartilage and proximal trachea.¹ Acquired laryngotracheal stenosis can result from a variety of causes, including iatrogenic injury, infections, systemic autoimmune disease, and idiopathic etiologies.^1,2 In the absence of a clear infectious or iatrogenic cause, the differential diagnosis commonly narrows to idiopathic and autoimmune etiologies. Distinguishing between these two entities is critical, as it directly informs clinical management.

Idiopathic subglottic stenosis (iSGS) accounts for approximately 20% of LTS cases and occurs in an estimated 1 in 400,000 individuals. It primarily affects Caucasian women between the third and fifth decades of life.¹ In contrast, the most common autoimmune cause of LTS is granulomatosis with polyangiitis (GPA),³ an antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis that classically involves the upper respiratory tract, lungs, and kidneys. GPA can affect all levels of the airway, with tracheal involvement reported in approximately 12–27% of patients.^4–7 Notably, airway disease may occur even in the absence of severe pulmonary or renal involvement,^4,5,8 and in some cases, may be the sole manifestation of GPA.⁹

Diagnosing GPA in patients with LTS can be challenging, particularly when extra-otolaryngologic features are absent. ANCA serologies can assist in distinguishing autoimmune from idiopathic disease, as ANCA positivity supports a diagnosis of GPA.³ However, ANCA testing lacks complete specificity and can be positive in several non-vasculitic conditions,¹⁰ and up to 25% of patients with LTS associated with GPA are ANCA negative.^4–8 Histopathology can help differentiate iSGS from GPA, but its diagnostic yield is extremely low. Biopsies from patients with tracheal involvement of GPA often show nonspecific inflammation, with diagnostic features such as granulomas, vasculitis, or necrosis being rare.^7,11,12

Clinical management differs between iSGS and GPA. In both conditions, serial endoscopic dilations and local corticosteroid injections are commonly employed, while cricotracheal resection or reconstruction may be considered for severe or refractory disease.^4,13–16 However, while GPA benefits from systemic immunosuppression in addition to procedural treatment,^8,12,17–19 immunosuppression is not commonly employed for treatment of iSGS and its potential benefit is still being explored.^20,21 Thus, distinguishing between idiopathic and autoimmune LTS is central to determining the appropriate therapeutic approach, particularly in assessing whether immunosuppression is warranted.

In this study, we describe a distinct pattern of laryngotracheal involvement—termed infracordal stenosis—to aid in this diagnostic distinction. This phenotype is characterized by swelling and stenosis originating at the inferior margin of the true vocal folds (TVFs), bounded by the conus elasticus. At times it presents along with the more typical subglottic stenosis that extends into the proximal trachea. In our clinical experience, infracordal stenosis is associated with autoimmune rather than idiopathic disease and may be uniquely responsive to glucocorticoids compared to typical SGS.

The objective of this study is to characterize infracordal stenosis as a distinct clinical phenotype within autoimmune LTS and to evaluate differences in clinical features and glucocorticoid responsiveness between patients with infracordal stenosis and typical AI-SGS. We conducted a retrospective analysis of patients with suspected autoimmune LTS seen at our medical center, focusing on a cohort of patients that were referred to rheumatology. At our center, patients are referred from otolaryngology to rheumatology when clinical suspicion for autoimmune disease exists based on atypical demographics for iSGS, positive ANCA serologies, extra-otolaryngologic manifestations, or inflammatory appearance of the airway. All patients with infracordal stenosis are referred to rheumatology out of concern for underlying autoimmune disease. The majority of these patients receive systemic treatment with prednisone, with or without steroid-sparing immunosuppression. In this context, we compared outcomes of patients with infracordal stenosis to those with typical AI-SGS. We hypothesized that patients with infracordal stenosis would be a similar demographic cohort to AI-SGS, have a greater responsiveness to glucocorticoid therapy, and thereby have a lower procedural burden compared with AI-SGS.

Materials and Methods

Patient Selection

IRB approval was obtained from the Johns Hopkins University Institutional Review Board prior to beginning this study. Informed consent was waived due to the retrospective nature of the study and minimal risk associated with data collection. Medical records were queried for all patients with ICD-10 codes associated with laryngotracheal stenosis (J38.6, J39.8, J98.09) who were evaluated by a rheumatologist between January 1, 2013 and December 2024 at Johns Hopkins Medical Institutions. Chart review was performed to confirm a diagnosis of LTS by an otolaryngologist. Patients were excluded if any of the following applied: (1) they were not seen by both the Rheumatology and Otolaryngology departments at Johns Hopkins, (2) there was no chart evidence of LTS, (3) they underwent dilation for LTS prior to presentation at Johns Hopkins, or (4) they had less than six months of follow-up after LTS diagnosis. We excluded patients who were diagnosed and treated for autoimmune LTS outside of our institution prior to presentation to standardize the time zero for time-to-event analysis.

Patients with infracordal stenosis were identified through a keyword search of laryngoscopy reports for the terms “infracordal” or “subcordal.” All flagged cases were reviewed by an otolaryngologist to confirm infracordal involvement, including review of laryngoscopic images when available. Infracordal stenosis was defined as airway narrowing at the level of the inferior TVFs, below the superior margin of the TVFs, and including—but not limited to—the space above the level of the cricoid cartilage (Figure 1).

Figure 1.

Image of the larynx showing anatomic locations of laryngotracheal stenosis. Infracordal stenosis is highlighted in red beginning at the level of the thyroid cartilage just inferior to the superior aspect of the true vocal folds contained by the conus elasticus.

Data Collection

We used a standardized data extraction form in REDCap to record data for all patients who met inclusion and exclusion criteria. We collected baseline data on the demographic characteristics and clinical characteristics of the cohort. Demographic data included age, sex, and race. Clinical characteristics included the date of LTS diagnosis, presence of airway symptoms (dyspnea, stridor, dysphonia, hemoptysis), anatomic location of airway stenoses, Cotton-Myer grades, and descriptors of gross appearance on endoscopy. For patients with a diagnosis of GPA, we collected additional disease-specific variables, including the date of GPA diagnosis, ANCA serologies (ANCA, anti-MPO, and anti-PR3 antibodies), and organ system involvement at diagnosis. Organ manifestations were coded using a structured list derived from the Birmingham Vasculitis Activity Score (BVAS) and the Vasculitis Damage Index (VDI).

Immunosuppressant medication data were abstracted by reviewing rheumatology and otolaryngology notes. We recorded whether patients were treated with prednisone at the time of laryngeal stenosis diagnosis as a binary variable; however, dose and duration were not recorded. We also documented start and stop dates of steroid-sparing immunosuppressants including rituximab, cyclophosphamide, methotrexate, azathioprine, leflunomide, and mycophenolate mofetil.

Procedural interventions were identified through review of clinical and procedural notes. For each patient, we recorded the dates and types of airway procedures, including balloon dilation, laryngotracheoplasty, laryngotracheal resection, tracheostomy, and stand-alone glucocorticoid injections. Adjunctive procedures performed at the time of dilation—such as local steroid injection or laser excision—were also recorded.

The follow-up period began at the time of LTS diagnosis. The primary outcome was time to first endoscopic dilation. Secondary outcomes included the dilation frequency over entire patient follow-up period, and percentage of patients with zero dilations.

Data Analysis

All analyses were conducted using R Version 4.5.0.²¹ Descriptive statistics were used to compare baseline characteristics between patients with and without infracordal stenosis. Categorical variables were compared using Fisher’s exact test, and continuous variables using t-tests. Time-to-event analysis was performed using Kaplan–Meier curves, with comparison of median time to first dilation between groups using the log-rank test. The number of dilations and dilation frequency over follow-up was compared using the Wilcoxon rank-sum test, due to the non-parametric distribution of the outcome.

Given the exploratory nature of this study and the limited sample size, all statistical analyses were performed for descriptive and hypothesis-generating purposes. Multiple clinical and treatment variables were compared between groups without formal adjustment for multiple comparisons. Accordingly, p-values should be interpreted cautiously and not as confirmatory evidence of statistical significance.

Results

A total of 49 patients with laryngotracheal stenosis were included in the study (Figure 2). Of these, 11 patients (22%) had infracordal stenosis, while 38 patients (78%) had typical AI-SGS. Among those with infracordal disease, 5 patients (45%) also had concurrent subglottic involvement, while 6 patients (55%) had isolated infracordal stenosis. Of the total 49 patients, 32 had available laryngoscopic images to confirm the presence or absence of infracordal and subglottic stenosis.

Figure 2. Patient selection flow diagram.

We queried medical records for patients with ICD-10 codes for laryngotracheal stenosis (J38.6, J39.8, J98.09) seen by a rheumatologist at Johns Hopkins between January 2013 and December 2024. Otolaryngologist-confirmed LTS was required. Patients with prior dilations were excluded to standardize time zero for time-to-event analyses. Patients with less than 6 months of follow-up were also excluded.

A summary of demographic, clinical, and treatment characteristics is provided in Table 1. GPA was diagnosed by a rheumatologist in 8 of 11 patients (73%) in the infracordal group and in 33 of 38 patients (87%) in the SGS group. ANCA positivity was observed in 67% of the overall cohort, with similar proportions across both groups. Among patients with GPA, there were no clear differences between groups in the presence of severe GPA disease manifestations.

Table 1.

Baseline demographic, clinical, and treatment characteristics, and outcome comparisons between laryngeal stenosis patients with and without infracordal involvement

	Infracordal Involvement n=11	No Infracordal Involvement n=38	p-value*
Demographics
Age, mean (sd)	44.5 (12.3)	42.6 (15.3)	0.69
Female, n (%)	7 (63.2)	26 (73.7)	0.71
Male, n (%)	4 (36.3)	10 (26.3)
Diagnosis
GPA, n (%)	8 (72.7)	33 (86.8)	0.50
Severe †	3 (27.3)	4 (10.5)	0.18
Non-severe	5 (45.5)	29 (76.3)
ANCA positive	7 (63.6)	26 (68.4)	0.81
Idiopathic SGS, n (%)	3 (27.3)	5 (13.2)
Cotton-Myer Grading
Grade I	4 (36.3)	17 (44.7)	0.18
Grade II	1 (9.1)	7 (18.4)
Grade III	5 (45.5)	5 (13.2)
Grade IV	0 (0.0)	0 (0.0)
Not reported	1 (9.1)	9 (23.7)
Airway Symptoms, n (%)
Dyspnea	9 (81.8)	29 (76.3)	1
Stridor	5 (45.5)	15 (39.5)	0.74
Cough	2 (18.2)	6 (15.8)	1
Dysphonia	5 (45.5)	12 (31.6)	0.48
Immunosuppression ‡, n (%)
Prednisone	10 (90.9)	27 (71.1)	0.41
Rituximab	3 (27.3)	13 (34.2)	1
Methotrexate	2 (18.2)	13 (34.2)	0.46
Azathioprine	3 (27.3)	4 (10.5)	0.12
Leflunomide	3 (27.2)	3 (7.9)	0.18
Cyclophosphamide	1 (9.1)	1 (2.6)	0.40
Mycophenolate mofetil	0 (0)	1 (2.6)	1

^*Statistical comparisons between groups were made using fisher’s exact tests, t-tests and wilcoxon rank sum tests where appropriate.

^†Severe disease defined by the presence of life or organ-threatening manifestations consistent with American College of Rheumatology/Vasculitis Foundation guidelines

^‡Medications received at the time of LTS diagnosis

Dyspnea was the most common presenting symptom across both groups. Dysphonia was reported in 45% of patients with infracordal stenosis compared to 31% of those with SGS. Cotton-Myer grades were reported in 39 participants. Patients with infracordal stenosis tended to have higher Cotton-Myer grades compared with those with SGS. At the time of airway stenosis diagnosis, 91% of patients with infracordal disease were treated with prednisone, compared to 71% of patients in the SGS group.

Patients with infracordal disease experienced longer time to first dilation compared to those without infracordal involvement. The median time to first dilation was 792 days in the infracordal group versus 44 days in the SGS group, as shown in Kaplan–Meier curves (Figure 3; p=0.048). No clear differences between groups were observed in dilation frequency or in the proportion of patients who required no dilations during the entire follow-up period (Table 2).

Figure 3.

Kaplan-Meier curves for time to first tracheal dilation in patients with infracordal stenosis versus typical subglottic stenosis. Time zero was the date of LTS diagnosis. Patients without dilation were censored at last follow-up or at two years. Comparison by log-rank test; p = 0.048.

Table 2.

Comparison of procedural outcomes between patients with and without infracordal involvement

	Infracordal Involvement (n=11)	No Infracordal Involvement (n=38)	p-value*
Follow-up time in years, mean (sd)	6.6 (4.3)	9.8 (5.7)	0.061
Time to first dilation in days, median (IQR)	729 days	44 days	0.048
Total dilations, median (IQR)	2.0 (0.0–3.0)	2.0 (1.0–4.5)	0.543
Dilations per year, median (IQR)	0.3 (0.0–0.4)	0.4 (0.1–0.7)	0.516
Patients with zero dilations, n (%)	4 (30.8)	8 (22.2)	0.708
Patients treated with in-office intra-lesional glucocorticoid injections, n (%)	1 (9.1)	8 (22.2)	0.758

^*Statistical comparisons between groups were made using wilcoxon rank sum test, and chi-square test where appropriate.

Among the six patients with isolated infracordal stenosis (without concomitant SGS), longitudinal disease trajectories were examined over the full duration of follow-up (Table 3). Four of six patients required no dilations; one of these four received a single series of intra-lesional glucocorticoid injections. Five of the six patients were treated with systemic glucocorticoids at presentation, all of whom reported symptomatic improvement at follow-up.

Table 3.

Descriptions of disease course for six patients with isolated infracordal involvement

	Follow-up Time (in years)	Time to first dilation (in years)	Number of Dilations	Intra-lesional steroid injections	Prednisone at diagnosis	Symptom improvement with GC
Patient 1	2.6	n/a	0	No	Yes	Yes
Patient 2	10.0	2.2	9	w/ dilation	No	n/a
Patient 3	5.8	3.5	2	w/ dilation	Yes	Yes
Patient 4	3.0	n/a	0	Yes	Yes	Yes
Patient 5	2.7	n/a	0	No	Yes	Yes
Patient 6	2.6	n/a	0	No	Yes	Yes

Discussion

In this study, we introduce a distinct phenotype of autoimmune LTS, termed infracordal stenosis, characterized by a unique clinical appearance and, based on our experience, increased responsiveness to glucocorticoid therapy compared to typical autoimmune SGS. Infracordal stenosis differs from typical AI-SGS both in location and endoscopic findings. Rather than the classic fibrotic stricture of iSGS, infracordal stenosis presents as edema involving the inferior aspect of the true vocal folds, at times with a connection along the posterior laryngeal wall (Figure 4, images d–f). In some cases, this infracordal involvement coexists with more typical AI-SGS. In this study, we evaluated clinical outcomes in 11 patients with infracordal stenosis and compared them to a matched group of patients with AI-SGS who had similar demographics, clinical and serologic profiles, and immunosuppressant exposure (Table 1). Patients with infracordal stenosis experienced a longer time to first dilation following diagnosis, and patients with isolated infracordal stenosis were often able to avoid procedural dilations entirely.

Figure 4.

Representative laryngoscopic images of subglottic stenosis versus infracordal stenosis.

Panel A: Laryngoscopy near view of subglottic stenosis with its typical appearance, starting inferior to the inferior aspect of the true vocal folds, with a fibrotic stricture circumferentially around the airway.

Panel B: Laryngoscopy far view of GPA-SGS with stenosis starting below the TVFs again with the typical stricture scar appearance.

Panel C: Laryngoscopy far view of iSGS with 80% stenosis distal to the inferior TVF. Note the similar appearance of subglottic stenosis to that of GPA-SGS.

Panel D: Laryngoscopy near view of infracordal stenosis showing the fullness and erythema of the inferior aspect of the true vocal folds. It begins inferior to the medial vibratory edge of the TVFs.

Panel E: Laryngoscopy far view infracordal stenosis starting immediately inferior to the medial edge of the TVFs. Additionally note the exudate in the inter-arytenoid area

Panel F: Laryngoscopy far view infracordal stenosis with swelling and erythema running into the inferior aspect of the TVF coming up from the subglottic area.

The observation of fewer procedural interventions among patients with infracordal stenosis raises important mechanistic and clinical considerations. We suspect that infracordal stenosis may be more responsive to systemic glucocorticoids compared to typical AI-SGS, accounting for this observation. All patients with isolated infracordal stenosis who received prednisone reported symptomatic improvement (Table 3), and representative pre- and post-treatment endoscopic images from one case (Figure 5) demonstrate marked resolution of edema following glucocorticoid therapy. These observations raise the possibility that infracordal stenosis represents an early or more inflammatory stage of autoimmune LTS, offering a therapeutic window during which systemic immunosuppression may reverse airway narrowing and avert procedural intervention. This interpretation is supported by the endoscopic appearance of infracordal stenosis, which is characterized predominantly by soft tissue edema rather than fibrotic narrowing (Figure 4). Future studies leveraging airway tissue analysis, including histopathology, may clarify whether infracordal stenosis reflects an early inflammatory phenotype or a distinct immunopathologic process within the spectrum of autoimmune LTS.

Figure 5.

Laryngoscopic images from a patient with infracordal stenosis before and after glucocorticoid treatment.

Panel A: Infracordal stenosis at presentation, prior to prednisone initiation, showing significant airway narrowing.

Panel B: Same patient after prednisone treatment, demonstrating continued infracordal stenosis but significantly increased glottic airway.

It is also conceivable that infracordal stenosis reflects a shared airway inflammatory pathway across multiple autoimmune disorders. Although most patients in our cohort carried a diagnosis of GPA, the potential relevance of this phenotype in other autoimmune or inflammatory airway diseases remains unexplored. Among patients with autoimmune LTS who do not meet full criteria for GPA, additional etiologies such as sarcoidosis, Behcet’s disease, or IgG4-related disease should be considered. Broader application of serologic, radiographic, and histopathologic evaluation in future studies may help delineate disease mechanisms and refine diagnostic classification within autoimmune laryngotracheal disease.

The demographic and clinical features of our cohort support that infracordal stenosis is more often associated with underlying autoimmune disease rather than idiopathic disease. In our series, 73% of patients with infracordal stenosis had a confirmed autoimmune diagnosis—a higher proportion than typically observed in patients with typical SGS. The observed male-to-female ratio of 1:2 also resembles that of GPA and contrasts with iSGS, which almost exclusively affects women. Moreover, the robust glucocorticoid responsiveness observed in all patients with isolated infracordal stenosis contrasts with the expected treatment response in iSGS. Collectively, these findings reinforce that infracordal stenosis represents a distinct, inflammation-predominant autoimmune airway phenotype.

Several limitations should be considered when interpreting the findings of our study. First, treatment practice patterns may have influenced outcomes. At our institution, patients with infracordal stenosis are often managed with greater confidence in glucocorticoid responsiveness, which may have led to fewer procedural interventions and contributed to the observed difference in dilation rates. Second, inherent anatomic differences between infracordal and subglottic stenosis could partly explain these results, as infracordal disease tends to involve the lateral airway walls, whereas subglottic stenosis is typically circumferential. Third, while we recorded whether patients were treated with glucocorticoids, because of logistical difficulties extracting data from electronic health records, we were limited in our ability to capture precise data on dose, duration, or adherence, limiting our ability to evaluate the relationship between treatment intensity and clinical outcomes.

Our relatively small sample size and further limited statistical power and precluded the multivariable modeling or advanced causal inference techniques. Given the small sample size, this study was not powered to detect modest differences between groups, and reported associations should be interpreted as hypothesis-generating rather than confirmatory. Future studies with larger cohorts and prospective data collection will be essential to validate these findings and better define the role of systemic immunosuppression in patients with infracordal stenosis. Such studies should incorporate the use of peak flow measurement to more precisely capture disease severity, treatment response, and disease progression.

Effective management of infracordal stenosis requires close, ongoing collaboration between otolaryngology and rheumatology. Rheumatologists play a critical role in evaluating for systemic autoimmune disease, initiating and monitoring immunosuppressive therapy, and assessing for extra-airway manifestations that may guide treatment intensity. Otolaryngologists are essential for diagnostic recognition through laryngoscopy, longitudinal airway surveillance, and procedural decision-making when interventions are required. Given the potential for infracordal stenosis to signal underlying autoimmune pathology—even in the absence of classic systemic features—early multidisciplinary involvement is essential to ensure timely diagnosis and to optimize outcomes.

Conclusion

Infracordal stenosis is a rare variant of laryngotracheal stenosis characterized by edema and narrowing of the inferior aspect of the true vocal folds, often extending to or below the level of the cricoid cartilage and sometimes coexisting with subglottic stenosis. In this study of patients with autoimmune LTS, those with infracordal involvement experienced a longer time to first dilation compared to those without. Our findings suggest that infracordal stenosis may represent a phenotype that is particularly responsive to systemic glucocorticoid therapy. Unlike AI-SGS, which often requires repeated procedural interventions, infracordal stenosis may be more effectively managed with systemic immunosuppression alone, potentially reducing the need for local procedures.